pinnacles and graduated panelling. It had, however, been lowered in
pitch and had been altered also by the insertion of a rather debased
Perpendicular window.[54] Whatever may be thought, therefore, of Sir
Gilbert Scott's action in rebuilding it, he has surely improved the
general effect of the front, and it is well that one of the roofs, at
any rate, should have been raised to the original pitch. What is most to
be regretted, perhaps, is the removal of all traces, if any there were,
of the chantry of the Holy Trinity _supra summum altare_, which was
situated, as its name implies, in the roof, behind the old gable.

In Archbishop Roger's day the choir was probably as long as it is now,
and Walbran (followed by Sir Gilbert Scott) believed that the aisles at
that period were returned across the east end. If so, the clearstorey
must have been a bay shorter than at present, with a pent roof
projecting from below it on the east side to cover the returned portion
of the aisle. The rebuilding of the east end in the Decorated period was
the first operation in which limestone was employed, but much of the old
gritstone has been used again.

[Illustration: THE NORTH-WESTERN PORTION OF THE NAVE.
(Junction of XIIth and XVIth Century work.)]

FOOTNOTES:

[33] The name Ripon comes probably from the Latin _ripa_, "a river's
bank." Bede uses a form "Inrhypum," which arose perhaps out of _in
ripis_. The derivation _Uri pons_ has been generally abandoned.

[34] The reason of the peculiarity here is the unusual width of the
nave. (_See below_, p. 44.)

[35] This will be explained in Chapter III.

[36] See illustration, p. 17.

[37] This was pointed out by Walbran.

[38] The Transitional or Transition-Norman work at Ripon probably was
not all erected during Roger's lifetime, but all of it will, in these
pages, be associated with his name.

[39] Upon a modern Chapter seal there is what is possibly meant for a
representation of Roger's church, with western towers, three spires, and
no aisles. The seal is a reproduction of another of the time of James
I., which may have been reproduced from a third of earlier date.

[40] For the origin and meaning of this knotwork, so often found in
these islands on ancient crosses, and for its value as an illustration
of the possible connection of Saxon architecture with the Comacine Guild
of Italy, see _The Cathedral Builders_, by Leader Scott, pp. 82-99, and
p. 145.

[41] This was th